A virtual machine might fail to power on if the .vmx file is changed in the backend.
Article ID: 409716
Updated On:
Products
Issue/Introduction
In certain scenarios, a virtual machine might fail to power on if its .vmx file or related files are deleted and recreated on an NFSv4.1 datastore while the ESXi host is still accessing the old file handle.
This issue is most often observed when restoring a VM from a backup or when external applications (for example, storage appliances such as Tintri VMstore) perform file operations on behalf of the VM.
As a result, ESXi encounters stale file handle errors and the VM cannot power on.
Error :
--------
An error was received from the ESX host while powering on VM <vm-name>.
Failed to open configuration file <vm-path>/<vm-name>.vmx.
vmkernel.log
------------------
WARNING: NFS41: FILE: 1259: stale file handle
WARNING: NFS41: NFS41FSDoOpenFile: Failed to open file <vm-name>.vmx: Stale NFS file handle
WARNING: NFS41: NFS41FSCompleteOpenFile: 3466: Failed to open <vm-path>/<vm-name>.vmx
vmkwarning.log
---------------------
Lost access to volume <datastore-name> due to stale file handle (NFSv4.1).
Environment
VMware vSphere ESXi 8.0 U2 and later.
NFSv4.1 datastores
Cause
Starting with ESXi 8.0 Update 2, Directory Name Lookup Cache (DNLC) was enabled for NFSv4.1.
If a file (such as the VM’s .vmx) is deleted and recreated while DNLC still caches the old handle, ESXi continues to use the stale file handle.
This leads to a stale handle error and prevents the VM from powering on.
Resolution
This issue is resolved in ESXi 8.0 U3g (8.0 P06).
Upgrade affected ESXi hosts to 8.0 U3g or later to apply the fix.
Feedback
